The Keynes-Townshend correspondence in 1937-38 makes it very clear that Keynes’s views on decision making, uncertainty, expectations and confidence in the General Theory are based directly on Keynes’s analysis of “non numerical” probabilities, Keynes’s own term for his interval valued probability based on approximation from Part II of the A Treatise on Probability, the weight of the evidence variable, w, from chapter 26 of the A Treatise on Probability, which is the mathematical version of the weight of the argument relation, V, from chapter 6 of the A Treatise on Probability, which established a logical relation between evidence statements in the form of propositions, and his decision weight approach based on his conventional coefficient of weight and risk, c. Keynes’s theories of effective demand, the demand for money, m.e.c., and liquidity preference explicitly incorporated expectations built on his conventional coefficient of weight and risk, c. The subjectivist approach of Irving Fisher, and especially the subjectivist, Bayesian, personalist view of Ramsey, is a special case of Keynes’s theory. Earlier work done by Fisher, Hawtrey, Robertson, Lavington, Pigou, and Knight , that also dealt with related views on expectations, effective demand, the demand for money, m.e.c., and liquidity preference, does not have any such explicit foundation in a logical theory of probability as put forth by Keynes in his A Treatise on Probability. Keynes’s definition of uncertainty in the General Theory, that uncertainty is an inverse function of weight, had nothing to do with the views of Knight or Shackle except that all three differentiate uncertainty for risk. The idea that Keynes’s General Theory was a grand synthesis that was put together with the ideas, work ,and analysis borrowed from many other economists, represents the same type of mistake made by Jevons, Viner, and Schumpeter concerning Adam Smith’s contributions in The Wealth of Nations. Jevons, Viner, and Schumpeter completely overlooked Smith’s original work in decision theory, his definition of uncertainty, and his general rejection of mathematical probability theory in favor of interval valued probability. The same kind of mistaken evaluation has taken place with regards to Keynes’s work in the General Theory. Keynes was the founder, originator, and developer of his own IS-LM model in chapters 13-15 of the General Theory. Keynes used his IS-LM model in Part IV of chapter 15 to demonstrate its superiority over the equation of exchange ,as well as to analyze the special case nature of classical and neoclassical economics. D. Champernowne’s 1936 article was the only review of the General Theory to correctly specify the arguments of the IS and LM curves, which included Q and Q’ variables to represent expectations, uncertainty and investor nervousness. Hicks removed these variables from his IS and LM curve in his article in Econometrica in 1937. The idea that, since Keynes was a Marshallian in his view of the appropriate role of mathematics in economics, he would work out his own four equation, simultaneous, IS LM model in December 1933, incorporate it into his 1934 draft of the General Theory, and then turn around and “burn the formal mathematical analysis” by leaving it out of the General Theory, is a fable that spread by the Robinson’s and R. Kahn.
Read full abstract